diff_of_means ratio_of_sd monthly_amplitude_ratio_of_means sign_correlation qqplot_mae acf_mae extremogram_mae
xgboost.mri_esm2_0.ssp245 -0.27% 0.985 0.837 0.519 4.216 0.098 0.036
xgboost.mri_esm2_0.ssp434 1.51% 0.985 0.862 0.510 3.237 0.087 0.038
xgboost.mri_esm2_0.ssp370 3.24% 0.978 0.865 0.508 6.473 0.066 0.033
xgboost.cesm2.ssp245 9.31% 0.912 0.913 0.515 18.596 0.011 0.025
xgboost.cesm2.ssp585 10.00% 0.896 0.886 0.525 19.994 0.014 0.035
xgboost.cesm2.ssp370 10.89% 0.889 0.873 0.500 21.765 0.015 0.034
cnn.cesm2.ssp585 15.94% 0.714 0.927 0.495 33.729 0.226 0.048
xgboost.ec_earth3.ssp434 18.53% 0.873 0.886 0.493 37.000 0.014 0.030
cnn.cesm2.ssp370 19.49% 0.743 1.000 0.503 39.221 0.256 0.041
cnn.cesm2.ssp245 20.25% 0.673 0.956 0.495 42.179 0.311 0.043
cnn.ec_earth3.ssp434 22.75% 0.753 0.907 0.483 45.576 0.185 0.045
nv.mri_esm2_0.ssp245 23.24% 0.987 0.839 0.510 46.388 0.137 0.042
nv.mri_esm2_0.ssp434 24.30% 0.988 0.869 0.514 48.509 0.134 0.039
cnn.mri_esm2_0.ssp370 24.39% 0.690 0.928 0.507 48.918 0.264 0.057
nv.mri_esm2_0.ssp370 26.97% 0.981 0.897 0.489 53.838 0.113 0.036
nv.cesm2.ssp245 28.08% 0.900 0.908 0.522 56.049 0.045 0.044
cnn.mri_esm2_0.ssp434 28.21% 0.650 0.910 0.511 56.341 0.287 0.057
nv.cesm2.ssp585 28.42% 0.898 0.932 0.518 56.725 0.040 0.039
cnn.mri_esm2_0.ssp245 28.57% 0.655 0.915 0.500 57.040 0.299 0.052
nv.cesm2.ssp370 29.28% 0.889 0.911 0.509 58.455 0.037 0.039
nv.ec_earth3.ssp434 35.05% 0.874 0.894 0.505 69.966 0.056 0.059

Time series of the first days

Distribution of daily values by month

QQ Plot

Distribution of the undownscaled value on days with estimated extremes values.

On the x-axis we have the daily mean (standardized). It says Undownscaled value, but is the daily mean after the downscaling. A good idea is to plot the original undownscaled value.

The purpose of this plot is to illustrate the distribution of P(undownscaled value | we predicted an extreme). This is useful because it reveals how much information we can recover concerning extreme events. If the distribution is skewed to the right, it suggests that we’re predicting extreme values only when extreme values have already occurred. Conversely, if the lower tail of the distribution resembles the reanalysis data, it indicates that we can capture short-duration extremes (e.g., brief periods of heavy rainfall, such as an intense downpour lasting an hour before stopping).

Autocorrelogram

Extremogram